criteriaST.cv: Cross-validation summaries

In geosptdb: Spatio-Temporal Radial Basis Functions with Distance-Based Methods (Optimization, Prediction and Cross Validation)

Cross-validation summaries

Description

Generate a data frame of statistical values associated with cross-validation

Usage

criteriaST.cv(m.cv)

Arguments

m.cv

data frame containing: prediction columns, prediction variance of cross-validation data points, observed values, residuals, zscore (residual divided by kriging standard error), and fold. If the rbfST.tcv function is used, the prediction variance, zscore (residual divided by standard error) will have NA's, coordinates data and time.

Value

data frame containing: mean prediction errors (MPE), average kriging standard error (AKSE), root-mean-square prediction errors (RMSPE), mean standardized prediction errors (MSPE), root-mean-square standardized prediction errors (RMSSPE), mean absolute percentage prediction errors (MAPPE), coefficient of correlation of the prediction errors (CCPE), coefficient of determination (R2) and squared coefficient of correlation of the prediction errors (pseudoR2)

Examples

# leave-one-out cross validation:
data(croatiadb)
coordinates(croatiadb) <- ~x+y

# inverse multiquadratic function, predefined eta and rho
tempm <- rbfST.tcv(MTEMP~X1+X2+X3+X4+X5+X6+X7+X8+X9+X10, croatiadb, eta=0.0108,
                   rho=0.00004, n.neigh=25, func="IM")
criteriaST.cv(tempm)

geosptdb documentation built on May 13, 2022, 1:05 a.m.